Obesity, which can lead to a variety of chronic diseases and reduced life span, continues to increase in prevalence in the US. A subset of obese subjects (30 percent of those entering medical weight-loss programs) has the recently characterized binge-eating disorder (BED), ingestion of very large meals without purging as occurs in bulimia nervosa. We have shown that stomach capacity is greater in obese than lean subjects, but this may be due mainly to the BED subset. We hypothesize that stomach capacity is larger in obese BED subjects than in obese non-BED subjects and lean subjects. Stomach capacity will be estimated by filling a gastric balloon in 24 obese BED and 24 non-obese subjects and 24 lean controls, each group divided equally by sex. Gastric capacity will be estimated from a) maximum volume tolerated and b) volume required to induce a given rise in intragastric pressure. Gastric capacity will be correlated with laboratory test-meal intake. Gastric emptying of a liquid meal will also be determined and related to gastric capacity. From out previous work, we would predict slower emptying, especially in the early phase, from a stomach with a large capacity. During the gastric emptying test, blood will be sampled for plasma cholecystokinin (CCK), a satiety peptide, shown to be lower in bulimics postprandially. The reduced CCK in bulimics may be the result of slower gastric emptying and larger gastric capacity and is predicted to be reduced in BED. The extent to which gastric capacity and emptying, test meal intake, and CCK change after restrictive dieting in obese subjects will then be determined. The obese BED and non-BED subjects will be randomized for 3 mo. of a) cognitive-behavior modification with a recommended intake of 5016 kJ/day (1200 kcal/day) or b) ad-lib intake. Gastric capacity may not decline in BED subjects if it is a primary factor in the disorder. Body composition, including underwater weighing wil be assessed before and after this period. Afterwards, a double-blind, dose-response study of a gastric emptying accelerant (cisapride) will be conducted in 12 non-dieting subjects of each group. Especially for the BED group, cisapride may speed emptying, enhance CCK release, and increase satiety. The lowest effective dose will then be used in a double-blind, 2 month crossover clinical trial of 12 obese non-BED subjects. Body weight and composition will be assessed at the start and monthly. The studies may enhance knowledge of the etiology, pathophysiology, and treatment of BED.